Pythium Insidiosum Keratitis: Histopathology and Rapid Novel Diagnostic Staining Technique

BASIC INVESTIGATION

Ruchi Mittal, MD, DNB,*† Shipra K. Jena, MSc,* Alisha Desai, MS,† and Sunil Agarwal, MD‡

Key Words: Pythiuim insidiosum keratitis, fungal keratitis, oomy- Purpose: To elucidate the histopathology of Pythium insidiosum cete, histopathology, iodine–potassium iodide–sulfuric acid stain keratitis and to describe a novel, simple, and rapid staining technique for identification of oomycete Pythium insidiosum and to differen- (Cornea 2017;36:1124–1132) tiate it from fungi. Methods: This is a laboratory investigation study of 38 non- ythium insidiosum, a pathogenic oomycete is the only consecutive cases (37 ocular samples and 1 colonic biopsy); 14 Preported agent to cause pythiosis in mammals, primarily microbiologically diagnosed as Pythium insidiosum keratitis and 24 causing disease in horses, dogs, and humans. Human as fungal keratitis. Review of clinical, demographic details, micro- pythiosis manifests itself as cutaneous, corneal, orbital, fi biological results, and identi cation of cases that necessitated vascular, gastrointestinal, and systemic forms of infection, evisceration was performed. Reevaluation of histopathology slides which can be devastating or life threatening.1–13 Pythium – was done using stains such as hematoxylin eosin, Gomori methe- infection is associated with high morbidity and mortality, and – – namine silver (GMS), periodic acid Schiff (PAS), potassium iodide its timely diagnosis and treatment is a chief cause for concern. sulfuric acid (IKI-H2SO4). Morphology, degree, and nature of Ocular pythiosis, commonly presents with corneal fl in ammation and load, distribution, and staining results of Pythium involvement and affects healthy individuals.5 The major insidiosum and its comparison with fungi were studied. challenge in management lies in the fact that it is misdiagnosed Results: and treated as a fungal infection, not only because of Delay in zoospore formation, failure of growth, and delay fi in identification of Pythium were the main cause of evisceration. insuf cient awareness among medical and laboratory specialists Corneal pythiosis showed epithelial ulceration, stromal destruction, and its morphological resemblance to fungi but also because of fl the lack of simple, quick, cost-effective, highly sensitive, and and varying in ammation; load and distribution of Pythium were fi inversely proportional to inflammation. The filaments were com- speci c diagnostic methods. Currently, there are no standard monly wide, with admixed narrower structures and uncommonly treatment protocols for pythiosis, and radical surgeries are involved Descemet membrane. The oomycete was not discretely advocated to completely excise the infected tissue to control the 8,14 discerned with PAS stain and stained distinctly with GMS stain and disease. Although oomycete Pythium insidiosum resembles IKI-H SO stain (100% sensitive). In comparison, fungal organisms fungi, under light microscopy it differs in its basic structural 2 4 composition and thus does not respond to antifungal medical stained well with PAS and GMS stain, but not with IKI-H2SO4 stain (100% specific). therapy. It contains cellulose in its cell wall, unlike fungi, which are chitinous and lack cellulose. The cell membrane of Pythium Conclusions: Pythium insidiosum keratitis is perhaps not more lacks ergosterol, unlike a fungal organism.15,16 devastating than fungal keratitis but late diagnosis, misdiagnosis, and It is imperative for all laboratory professionals to be treatment as fungal infection are major heralds. Early diagnosis may aware of the morphology and distinctive identification markedly improve the patient outcome. IKI-H2SO4 is a cost- features of this pathogenic oomycete. Zoospore formation, effective, simple, sensitive, and specific stain for the diagnosis of immunohistochemical detection, noncommercial serological oomycete Pythium. tests, and gene sequencing techniques have been studied for its diagnosis and differentiation from fungal organisms.5,6,17–23 Although zoospore formation is recommended as an important Received for publication February 6, 2017; revision received April 4, 2017; fi 5,17 accepted April 9, 2017. Published online ahead of print June 2, 2017. diagnostic tool for identi cation of Pythium, it is known to From the *Dalmia Ophthalmic Pathology Services, L. V. Prasad Eye Institute, be fraught with limitations. In addition to a delay period of 3 Bhubaneswar, India; †Tej Kohli Cornea Institute, L. V. Prasad Eye to 7 days, absence of growth in media in more than 30% of Institute, Bhubaneswar, India; and ‡Department of Pathology, Kalinga cases5 can limit its identification by zoospore formation or Hospital Limited, Bhubaneswar, India. molecular diagnostic tools. In addition, zoospore formation Supported by Hyderabad Eye Research Foundation and Tej Kohli Cornea Institute. cannot be demonstrated if the biopsy material is sent to the The authors have no conflicts of interest to disclose. histopathology laboratory alone. S. K. Jena has equal contribution as first author. In such a scenario, it is imperative that professionals are Reprints: Ruchi Mittal, MD, DNB, Dalmia Ophthalmic Pathology Services adequately trained and equipped to accurately identify and Tej Kohli Cornea Institute, L. V. Prasad Eye Institute, SMTC Campus, Patia, Bhubaneswar 751024, Orissa, India (e-mail: dr.rmittal@ Pythium from all laboratory samples, including surgically gmail.com). resected formalin-fixed paraffin-embedded (FFPE) tissues; Copyright © 2017 Wolters Kluwer Health, Inc. All rights reserved. hence, we performed a study to identify oomycete Pythium

1124 | www.corneajrnl.com Cornea Volume 36, Number 9, September 2017

Copyright Ó 2017 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited. Cornea Volume 36, Number 9, September 2017 Pythium Insidiosum Keratitis insidiosum using simple staining techniques and also to study (PA). All the slides were treated in PA (0.5% and 1.0%) in the morphological features of Pythium insidiosum keratitis. various batches for 0.5, 2, 3, and 5 minutes. Schiff reagent was used without any modifications for 10 minutes. The difference MATERIALS AND METHODS in the intensity of staining was noted as negative, 1+ to 4+ with different concentrations and durations of staining. Inability to The study was conducted at L. V. Prasad Eye Institute, visualize the filamentous structures was recorded as negative. India. Informed consent was obtained from all patients, and the When the structures appeared pink, similar to stromal staining, study was approved by the institutional review board, reference it was called as 1+. Intensity of staining similar to Descemet 2016-87-IM-15. The corneal scrapings of patients with clinically membrane (DM) was called 4+. suspected microbial keratitis, processed in the microbiology department as per our institute protocol,24 were included. Thirty- seven nonconsecutive blocks of FFPE ocular tissues [32 half Validation of IKI-H2SO4 Staining Techniques corneal buttons (CBs) and 5 evisceration specimens] of 36 on Corneal Scrapes and Pure Cultures patients were collected, for which the species of the infective In addition, slides of 7 corneal scrapings stained pre- organism was already identified and confirmed in the microbi- viously with potassium hydroxide-calcofluor white (KOH- ology department. Fourteen of these FFPE tissues (10 CBs and 4 CFW) and Gram stain and diagnosed as having either fungi eviscerations) of 13 patients had microbiological diagnoses of or Pythium (referred as SCR 01–07 on morphology alone) were Pythium insidiosum, based on the morphology and zoospore also retrieved. These slides were destained, hydrated, and formation, and confirmed by ITS DNA sequencing. Twenty- subjected to IKI solution (iodine concentration: 0.4 g) for 2 three of the tissues (22 CBs and 1 evisceration) were cases of hours followed by 40 mLof65%HSO and observed under fungal keratitis and served as controls. These included Asper- 2 4 a microscope after placing a cover slip. These cases were also gillus spp.(10),Fusarium spp.(8),Cladosporium (2), and one studied for growth in culture media [blood agar, chocolate agar, each of Acremonium, Candida,andColletotrichum.Thecases and Sabouraud dextrose agar (SDA)]. Samples that did not of fungal keratitis that had resolved with medical therapy were grow on SDA but showed typical colonies on chocolate agar or not included. Gastrointestinal tissue of one case of Mucormy- blood agar were studied for zoospore formation in the cosis was also included in the study. Adequacy and the presence induction medium, using carnation leaf. Smears of pure of the infective organism in all the paraffinblockswere cultures (from SDA/blood agar/chocolate agar) of these 7 confirmed using hematoxylin–eosin (H&E) stain and Gomori cases were also studied by staining with IKI solution for 2 methenamine silver (GMS) stain, and all the slides were re- hours and 65% H SO . viewed by 2 senior pathologists (R.M. and S.A.). Sections for 2 4 staining were deparaffinized by heating at 55 to 60°C for a duration of 20 minutes followed by immersing in xylene (15 Validation of the Staining Technique on dips)-3 changes, subsequently rehydrated using graded alcohol Archived Destained H&E Slides and washing in water. Iodine–potassium iodide–sulfuric acid Archived slides of 10/13 cases of microbiologically stain (IKI-H SO ) and periodic acid–Schiff (PAS) stain with 2 4 diagnosed Pythium keratitis, previously stained with H&E modifications were performed to differentiate between Pythium stain were retrieved and stained with IKI-H SO stain. insidiosum keratitis and fungal keratitis. 2 4 Clinical record sheets/electronic medical records of all the cases were reviewed, and demographic and clinical Iodine–Potassium Iodide (IKI)—Sulfuric Acid information was collected. This included age/sex, occupation, Stain (IKI-H2SO4) history of trauma, duration of symptoms, duration between IKI solution was prepared in an amber-colored bottle by presentation to therapeutic penetrating keratoplasty (TPK), dissolving 2 g of potassium iodide (KI) in 100 mL of distilled and management of infection. water and adding 0.2 g of iodine to it. This mixture was then Statistical analysis: Detection sensitivity and specificity allowed to stand overnight for complete dissolution of iodine in were calculated using Microsoft Excel 2013 software. the solution. Deparaffinized and rehydrated sections on glass slides were placed in this solution for a duration of 45 minutes and as a separate batch for 2 hours. These slides were then RESULTS removed, and excess stain was wiped off carefully with an The mean age of presentation was 43 years (age range: absorbent (tissue) paper. Then, 40 mL of 65% sulfuric acid 28–76 yrs) and the male:female ratio was 3:1. Four of 13 (H2SO4) was carefully placed over the slide, cover slipped, and patients with Pythium keratitis were farmers by occupation, with was immediately observed under the bright field microscope. a history of trauma in 7/13. The duration of symptoms at the Observation of change of color of hyphal/filamentous structures time of presentation of Pythium keratitis varied from 4 to 45 to bluish/bluish black was reported as positive and yellowish to days with a mean of 16 days, compared with 24 days for yellowish brown was reported as negative staining. Fusarium and Aspergillus spp. keratitis. Patients presented with symptoms of pain, redness, irritation, and photophobia. Best- corrected visual acuity varied from 20/400 to hand movements. PAS Stain The corneal infiltrates resembled a fungal ulcer with typical PAS stain was performed with modifications in the feathery margins. In addition, a characteristic reticular pattern of concentration and duration of treatment with periodic acid subepithelial and superficial stromal infiltrates was noted.

Copyright Ó 2017 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited. Mittal et al Cornea Volume 36, Number 9, September 2017

Distinctive dot-like infiltrates were seen emanating from the Hematoxylin–Eosin Stain for Pythium Cases central infiltrate (Fig. 1). Perforation and hypopyon were noted The corneal tissue showed epithelial ulceration, with in 5/13 patients at presentation. All 13 patients were clinically patchy stromal destruction, thinning, and infiltrates in 5/10 diagnosed with fungal keratitisandwerereportedtohave cases; and diffuse stromal destruction (Fig. 2A), ulceration, and aseptate or poorly septate broad filamentous structures, compat- infiltrates in the rest of the 5 cases. Patchy infiltrates were ible with fungal keratitis on KOH mount. They were treated with predominantly present in paracentral and peripheral corneal natamycin and ketoconazole, with no response to medical stroma, either anteriorly and/or posteriorly (Fig. 2B, marked therapy. Ten of thirteen patients underwent TPK; the time with open parenthesis). Stromal infiltrates varied from mild interval between initial presentations to TPK varied from 1 to 36 (Fig. 2B) to severe,25 primarily comprised of polymorphonu- days, with a mean of 11.6 days for Pythium cases and 10 days clear leukocytes (PMNL), admixed with nuclear debris and for Fusarium spp. and Aspergillus spp. Three of 13 patients of eosinophils (Fig. 2C); distribution and severity of the infiltrate Pythium keratitis underwent evisceration; the median time was distinctly inversely proportional to the load of Pythium interval between initial presentation to evisceration was 36 days (Fig. 2D, marked with open parenthesis). Infiltrates were seen (26–44 d). Two of these 3 patients had to undergo evisceration reaching the edge of TPK in 7/10 cases. The TPK, in 5 cases, for endophthalmitis because of delay in zoospore formation was large enough to include limbal conjunctiva, which showed and the other because of failure of growth in culture media. One markedly dense infiltration of PMNLs, eosinophils, lympho- of 10 patients who underwent TPK presented with graft cytes, and plasma cells (Fig. 2E). Granulomatous inflammation recurrence after 15 days and subsequently needed evisceration. was seen in 3/10 cases, exuberant in 1/10 (Fig. 2F, PAS), primarily noticed along the posterior stroma and fragmented DM, comprising multinucleated giant cells and histiocytes. DM could be discerned in 7/10 cases, continuous in 3/7, and fragmented in 4/7 cases. There was complete absence of endothelial cells in all the corneal sections; anterior chamber (AC) exudates were seen in 5/10 cases, varying from mild to severely dense infiltrates, comprising PMNLs, eosinophils, and histiocytes amidst debris material. On close observation, the filaments of Pythium appeared pale pinkish ghost-like structures on H&E stain, with intensity similar to stromal staining.

GMS Stain for Pythium Cases The filaments of Pythium stained brown and stroma stained greenish. The staining intensity was similar to fungi. The morphology of Pythium filaments could be best studied on GMS stain and these appeared as broad (Fig. 3A, thick arrow) to narrow (Fig. 3A, thin arrow), long to short, aseptate to septate (Fig. 3A, elbow arrow), without or with perpendicular to obtuse angle branching (Fig. 3A, curved arrow). Interspersed filaments showed hyphal swellings (Fig. 3A, inset, arrow head). They had admixed filaments, which appeared as collapsed, twisted, folded hollow tubes, or cut ends (end on view). The filament load was heavy in 6/10 cases (Fig. 3B), moderate in 1/10, and low in 3/10. The Bowman membrane was seen to be involved in one case (Fig. 3B, marked with arrows). The case with severe granulomatous inflammation showed mild load of organisms, some within the giant cells (Fig. 3C, marked with arrows). The organism was seen reaching the resection margin of the TPK in 2cases.Thefilaments concentrated in the posterior stroma, pre- DM (Fig. 3D), indenting DM; however, it was intriguing that only an occasional filament was seen involving DM in only 1/10 cases(Fig.3E,markedwithanarrow).Asinglefilament was noted in the AC in 1/10 cases. The adhered uveal tissue/uveal vessels did not show any filamentous structures.

Histopathology of Evisceration Specimens of FIGURE 1. Slit-lamp photograph (A) showing subepithelial Pythium and superficial stromal infiltrates in a distinctive “reticular” Cases pattern and (B) diffuse dot-like infiltrates emanating from the The evisceration specimens (4) showed diffuse main infiltrate extending to the peripheral cornea. corneal stromal destruction, an inflammatory mass

Copyright Ó 2017 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited. Cornea Volume 36, Number 9, September 2017 Pythium Insidiosum Keratitis

FIGURE 2. Histopathology of Pythi- um insidiosum keratitis. A, Scanner magnification shows a perforated corneal ulcer with diffuse inflammatory infiltrates [hematoxylin–eosin (H&E), original magnification ·20]. B, Superficial ulceration with a mild degree of predominantly anterior and posterior stromal inflammatory infiltrate (·100, marked with open parenthesis) and (C) shows predominantly PMNLs admixed with eosinophils and fibrinous debris (·600). D, There is almost full-thickness and diffuse involvement by the filaments of Pythium (GMS stain, original magnification ·100, marked with orange open parenthesis). E, Limbal conjunctiva shows predominantly PMNLs, eosinophils in a loose stroma with thin-walled capillary channels (H&E, original magnification ·600). F, Posterior corneal stroma shows predominantly granulomatous inflammation, comprising of multinucleated giant cells (diamond shape marked) with en- gulfed ghost-like Pythium filament (marked with arrows), fragmented DM is noted with adhered uveal pigment (PAS, original magnification ·600). comprising stromal tissue, AC exudates, iris, ciliary body PMNLs were the principal inflammatory cell; eosinophils (Fig. 3F), and anterior lens capsule. The inflammation were also seen in all the cases. Granulomatous inflamma- comprised PMNLs, admixed with eosinophils, plasma tion was seen in 15% of cases. The filaments stained cells, and debris material. Few histiocytes and multinu- pinkish with bluish hue, some appeared ghost-like; cleated giant cells were seen in the posterior stromal however, some internal structures of the fungi also stained. exudates and along the DM. Uveal stroma showed the presence of chronic lymphomononuclear cells. The retina and vitreous also showed the presence of eosinophils and GMS Stain for Non-Pythium Cases PMNLs. The load of Pythium was heavy in the corneal (Fungal Keratitis) stroma, and DM was seen infiltrated by Pythium in only 1/ All the fungi stained brownish, and stroma appeared 4 cases; the rest of the cases showed filaments just anterior greenish. The fungal filaments were seen reaching the to DM. The inflammatory mass in the AC showed an posterior stroma and involving DM in 89% of cases. occasional filament in one case; the lens capsule, uveal tissue, uveal vessels, retina, and vitreous did not show any filamentous structures. Analyses of PAS and Modified PAS Stain for Pythium Cases Pythium filaments failed to stain with PAS Hematoxylin–Eosin Stain for Non-Pythium stain, when treated with PA for a duration of less than Cases (Fungal Keratitis) 3 minutes, or with a concentration of 0.5%. After treatment The fungal load varied from moderate to severe, with PA (1%) for 5 minutes, the filaments were not inversely proportional to the extent of inflammation. discretely identifiable and stained pale pinkish (1+, Fig.

Copyright Ó 2017 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited. Mittal et al Cornea Volume 36, Number 9, September 2017

FIGURE 3. Morphology of Pythium insidiosum (A) broad filaments (thick arrows), narrow filaments (thin arrow), septations (elbow arrow), and branching (curved arrows); the inset shows a globular swelling of the filament (original magnification ·600, arrow head, GMS). B, Heavy load of Pythium with involvement of the Bowman membrane (marked with arrows); (C) the presence of the infective organism within the cyto- plasm of giant cells is noted (marked with arrows, original magnification ·600). D, Shows predominance of filaments in the posterior stroma, reaching the Descemet and indenting it; however, no filaments were seen within Descemet and in the AC exudates (original magnification ·600). E, Numerous filaments are seen along the pre-DM, and an occasional filament can be observed within the structure of DM (original magnification ·600, marked with an arrow, GMS). F, A perforated corneal ulcer is observed with AC exudates, and uveal tissue with a marked degree of inflammatory infiltrates (H and E ·100).

4A; curved arrows) to pale magenta (2+, Fig. 4A; arrows). Analysis of IKI-H2SO4 Stain for Non-Pythium Pythium sections never stained at an intensity of 3+/4+, Cases (Fungal Keratitis) irrespective of the concentration and staining time. None of the fungal organisms stained bluish to black with IKI-H2SO4 stain, they appeared yellowish to brown and were difficult to discern as the stroma also stained similarly (Fig. 4F). Analysis of PAS and Modified PAS Stain for Fibro-keratocytic cells or inflammatory cells/debris did not show Non-Pythium Cases (Fungal Keratitis) a change to bluish/bluish black structures. A well-prepared PAS stain solution stained the fungal filaments at an intensity of 3+ to 4+ at the end of 3 to 5 minutes at 0.5% to 1% concentration of PA (Figs. Results of Staining of Scrape Smears 4B, D–F). and Culture Followed by standardization of staining methods on half CBs, staining was performed on 7 direct corneal scrapes (Figs. Analysis of IKI-H2SO4 Stain for Pythium Cases 5A–F). Two of 7 scrapes showed bluish staining of the All 14 tissues (13 cases) showed blue staining of the filaments with IKI-H2SO4 solution (SCR 01 and SCR 05, filaments (Figs. 4C–E) at the end of 45 minutes; stroma Figs. 5A, B). Samples of SCR 01 also grew white feathery appearing yellow to yellowish brown. The stroma started colonies in blood and chocolate agar; these colonies stained to crumble and break as soon as the acid came in contact blue on treatment with IKI-H2SO4 solution (Fig. 5C). At the with the stroma, but still remained intact enough to end of day 7, this case also demonstrated zoospore formation. allow observations. No other ocular structure showed However, SCR 05 did not grow in the culture medium, bluish staining. therefore culture isolate was not available to stain with IKI-

Copyright Ó 2017 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited. Cornea Volume 36, Number 9, September 2017 Pythium Insidiosum Keratitis

FIGURE 4. A and B, PAS stain results (1% PA, 5 min). A, Filaments of Pythium with poor staining (1+ to 2+; original magnification ·600). B, Dis- tinct staining of fungal filaments (4 +). C–E, Iodine–potassium iodide– sulfuric acid (IKI-H2SO4) stains results for Pythium cases. The filaments of Pythium are highlighted, staining blue (original magnification ·400), one filament can be observed cross- ing through DM to reach the AC (marked with arrow). F, IKI-H2SO4 stain results for fungal keratitis: Fun- gal filaments appear yellowish brown, the corneal stroma also ap- pears yellowish brown (original magnification ·100).

H2SO4 stain. The rest of the scrapes were found to have fungal stain or stains variably, from pinkish (1+) to pale magenta filaments, which grew on SDA and neither the direct scrape (2+). This poor or no staining of Pythium with PAS stain smears (SCR 02–04, 06, 07; Figs. 5D, E) nor the smears of provides a clue of a nonfungal or a fungal-like organism for culture isolate (Fig. 5F) stained blue with IKI-H2SO4 solution. a keen observer, when an internal or an external positive control is available for comparison (DM readily serves as an internal positive control). In fact, it was this poor staining of Results of Staining of Destained these apparently fungal organisms, on PAS stain, that Archived Slides prompted the authors to study the staining, using different IKI-H2SO4 stain demonstrated the filaments of Pythium concentrations and timing of exposure to PA. Later, after we distinctly as blue, results exactly similar to FFPE sections that identified these as Pythium, the difference between staining were not destained. reaction of Pythium and fungi could be easily explained, based on the difference in the composition of their cell wall. Cellulose gives a weaker reaction than chitin and pectin, so it DISCUSSION needs a longer time of exposure to PA for complete oxidation Definitive diagnosis of Pythium insidiosum infection is to aldehydes, which then reacts with Schiff reagent to give always dependant on laboratory evaluation. The organism is a magenta color.27 We observed that not only the duration but seen as broad, branching, aseptate to sparsely septate also the concentration of PA also showed a difference filaments, and thus mimics Zygomycosis.5,26 These structures in staining. stain like fungi on GMS stain. PAS stains fungi with bright The chemistry of the cell wall of Pythium further magenta color; however, Pythium insidiosum either does not prompted us to develop a simple, quick, inexpensive stain for

Copyright Ó 2017 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited. Mittal et al Cornea Volume 36, Number 9, September 2017

FIGURE 5. Photomicrograph of the corneal scrapings. A–C, Pythium previously stained with KOH-CFW (A) or Gram stain (B), when destained and restained with IKI-H2SO4 stains the filamentous structures bluish to blackish. Pure culture isolate also showed blue staining of the filaments (C). Non-Pythium scrape smears (D–F), when destained and restained with IKI-H2SO4 stained the fungal filaments yellowish to yellowish brown (D–E). F, Culture isolate of fungal colony also stained yellowish brown in color (original magnification ·400).

Pythium. When cellulose is treated with acids, the crystalline in identification/failure of growth in culture media was the structure of cellulose is destroyed with formation of oligo- major reason for the cases of Pythium keratitis to progress to saccharides and sugars, which is preceded by formation of an endophthalmitis, this stain would be helpful for immediate intermediate complex known as hydrocellulose, which can identification of the infective organism, thus would allow form a blue adsorption compound with iodine solution.28 This planning of definitive management of the patient at their very acid hydrolysis reaction rate may vary with the concentration first visit, relieving the patient from misdiagnosis of fungal of acid, temperature of the reaction mixture, and of the infection and an extra burden of initial treatment with environment. Using the above reaction principle, we demon- antifungal therapy. Another advantage is that in situations strated that oomycete Pythium can thus be differentiated from in which the scrape material is scant, the slides initially fungal organisms. Our study showed excellent results not stained with KOH and gram stain can be destained and only on FFPE tissue sections but also on archived destained restained to demonstrate excellent results with IKI-H2SO4 H&E stained slides, fresh scrape smears, destained KOH/ stain. None of the sections (23 cases), direct scrapes (5), and CFW, and Gram-stained slides. During staining and interpre- culture isolates (5) of fungal keratitis demonstrated bluish tation, care is advocated, because use of absorbent tissue filaments, which assured a 100% specificity of the stain paper, filter paper fibers, cotton, and vegetable fibers in for oomycetes. colonoscopic biopsies will also give a positive reaction. It was interesting to note that the host reaction to However, laboratory specialists are well aware of fungal Pythium was almost similar to that to fungal keratitis; most morphology and are trained to identify artifacts. conspicuous was the inverse correlation between the degree of Identification of a case as Pythium was based on the inflammatory infiltrates and load of Pythium. Granulomatous 29–31 criteria previously published in the literature. IKI-H2SO4 inflammation was observed in 75% of evisceration specimens stain showed 100% sensitivity in detection of the organism, and 30% of TPK specimens of Pythium,comparedwith and the test could be performed within 2 hours of collection reported 14% of TPK specimens of fungal keratitis. It was of the scrape material. Because our study showed that a delay intriguing to observe that Pythium, unlike fungi, was not

Copyright Ó 2017 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited. Cornea Volume 36, Number 9, September 2017 Pythium Insidiosum Keratitis commonly seen involving/growing within the substance of REFERENCES DM, but uncommonly invading it through micropunctures in 1. Hilton RE, Tepedino K, Glenn CJ, et al. Swamp cancer: a case of human DM to reach the AC. In addition, although Pythium is known pythiosis and review of the literature. Br J Dermatol. 2016;175:394–397. to cause life-threatening vascular infections, it was fortunately 2. Krajaejun T, Pracharktam R, Wongwaisayawan S, et al. Ocular pythiosis: is it under-diagnosed? Am J Ophthalmol. 2004;137:370–372. seen to spare vessels of the eye/uvea. It was also not seen to 3. Thanathanee O, Enkvetchakul O, Rangsin R, et al. Outbreak of Pythium involve any other ocular structures, even in patients who keratitis during rainy season: a case series. Cornea. 2013;32:199–204. progressed to endophthalmitis. This could perhaps imply that 4. Mendoza L, Hernandez F, Ajello L. Life cycle of the human and animal Pythium commonly does not invade other ocular structures, but oomycete pathogen Pythium insidiosum. J Clin Microbiol. 1993;31: fl 2967–2973. it is the in ammatory cascade that may be the major role player 5. Krajaejun T, Sathapatayavongs B, Pracharktam R, et al. Clinical and in pathogenesis of ocular pythiosis. Although a larger study epidemiological analyses of human pythiosis in Thailand. Clin Infect Dis. with more patients is needed to confirm the morphological 2006;43:569–576. observations and understand the pathogenesis, an early diag- 6. Inkomlue R, Larbcharoensub N, Karnsombut P, et al. Development of an nosis and targeted management/surgery may definitely prevent anti-Elicitin antibody-based immunohistochemical assay for diagnosis of pythiosis. J Clin Microbiol. 2016;54:43–48. evisceration/enucleation and improve visual prognosis. 7. Kirzhner M, Arnold SR, Lyle C, et al. Pythium insidiosum: a rare Our study also observed that the filamentous structures necrotizing orbital and facial infection. J Pediatr Infect Dis Soc. 2015;4: of Pythium insidiosum may not necessarily be only broad e10–e13. and aseptate, but may also be comparatively narrower and 8. Bosco Sde M, Bagagli E, Araújo JP Jr, et al. Human pythiosis, Brazil. – show septations. These narrow filaments may incidentally be Emerg Infect Dis. 2005;11:715 718. 9. Farmer AR, Murray CK, Driscoll IR, et al. Combat-related pythium the only structures observed on corneal scrapes, and in such aphanidermatum invasive wound infection: case report and discussion of a scenario, it is still prudent to exclude Pythium as the sole utility of molecular diagnostics. J Clin Microbiol. 2015;53:1968–1975. etiological agent, or there may be a mixed fungal and 10. Franco DM, Aronson JF, Hawkins HK, et al. Systemic Pythium Pythium infection. The main filaments of Pythium vary from insidiosum in a pediatric burn patient. Burns. 2010;36:e68–e71. 4to6mm in thickness and their lateral branches are thinner, 11. Sathapatayavongs B, Leelachaikul P, Prachaktam R, et al. Human m fi pythiosis associated with thalassemia hemoglobinopathy syndrome. J measuring about 2.5 m. The septations in the laments Infect Dis. 1989;159:274–280. depend on the age of the hyphal structures and nature of 12. Imwidthaya P. Human pythiosis in Thailand. Postgrad Med J. 1994;70: culture media.32 558–560. One minor limitation of this staining technique is that it 13. Gaastra W, Lipman LJ, De Cock AW, et al. Pythium insidiosum: an overview. Vet Microbiol. 2010;146:1–16. requires handling of concentrated H2SO4 (65%); therefore, 14. Permpalung N, Worasilchai N, Plongla R, et al. Treatment outcomes of caution is recommended. The slides once evaluated cannot be surgery, antifungal therapy and immunotherapy in ocular and vascular permanently archived; however, microphotographs can be human pythiosis: a retrospective study of 18 patients. J Antimicrob captured during stain evaluation and archived for Chemother. 2015;70:1885–1892. future reference. 15. Alexopoulos CJ, Mims CW, Blackwell M. Introductory Mycology. John Wiley & Sons: New York; 1996. To conclude, IKI-H2SO4 can be used as a sensitive and 16. Kamoun S. Molecular genetics of pathogenic oomycetes. Eukaryot Cell. a very specific stain for identification of pathogenic oomycete 2003;2:191–199. Pythium insidiosum. We would recommend that all scrape 17. Sharma S, Balne PK, Motukupally SR, et al. Pythium insidiosum smears of clinically suspected fungal keratitis should also be keratitis: clinical profile and role of DNA sequencing and zoospore formation in diagnosis. Cornea. 2015;34:438–442. evaluated by IKI-H2SO4 stain because it can help in early fi 18. Keeratijarut A, Karnsombut P, Aroonroch R, et al. Evaluation of an in- identi cation of Pythium. Clinicians may suspect Pythium by house immunoperoxidase staining assay for histodiagnosis of human the characteristic reticular pattern of subepithelial and anterior pythiosis. Southeast Asian J Trop Med Public Health. 2009;40:1298–1305. stromal infiltrates, diffuse dot-like infiltrates emanating from 19. Keeratijarut A, Lohnoo T, Yingyong W, et al. A peptide ELISA to detect the central infiltrates, and pathologists by the morphology of antibodies against Pythium insidiosum based on predicted antigenic filaments, their concentration along the posterior stroma with determinants of exo-1,3-beta-glucanase. Southeast Asian J Trop Med Public Health. 2013;44:672–680. uncommon involvement of DM, poor staining results on PAS 20. Mendoza L, Kaufman L, Mandy W, et al. Serodiagnosis of human and stain and confirmed on IKI stain. Ideally, IKI-H2SO4 stain animal pythiosis using an enzyme-linked immunosorbent assay. Clin should be included in the battery of stains used for diagnosis Diagn Lab Immunol. 1997;4:715–718. and differentiation of fungal and fungal-like infections 21. Intaramat A, Sornprachum T, Chantrathonkul B, et al. Protein A/G-based (oomycete), which may be missed if only PAS stain is immunochromatographic test for serodiagnosis of pythiosis in human and animal subjects from Asia and Americas. Med Mycol. 2016;54:641–647. performed or misdiagnosed as fungal infection with GMS 22. Jindayok T, Piromsontikorn S, Srimuang S, et al. Hemagglutination test stain. Early and accurate diagnosis of this underrecognized/ for rapid serodiagnosis of human pythiosis. Clin Vaccin Immunol. 2009; undiagnosed/misdiagnosed infection will help us to better 16:1047–1051. understand its pathophysiology and will also pave our path to 23. Grooters AM, Gee MK. Development of a nested polymerase chain reaction assay for the detection and identification of Pythium insidiosum. develop more effective medical therapy for this vision- J Vet Intern Med. 2002;16:147–152. threatening/life-threatening infectious disease. 24. Gopinathan U, Sharma S, Garg P, et al. Review of epidemiological features, microbiological diagnosis and treatment outcome of microbial keratitis: experience of over a decade. Indian J Ophthalmol. 2009;57: ACKNOWLEDGMENT 273–279. 25. Vemuganti GK, Garg P, Gopinathan U, et al. Evaluation of agent and Authors wish to express their gratitude towards Manas host factors in progression of mycotic keratitis: A histologic and Ranjan Barik, Aparajita Mallick, Kalandi Charan Muduli microbiologic study of 167 corneal buttons. Ophthalmology. 2002;109: and Shilpa Priyadarshini Rout for their technical help. 1538–1546.

Copyright Ó 2017 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited. Mittal et al Cornea Volume 36, Number 9, September 2017

26. Mendoza L, Prasla SH, Ajello L. Orbital pythiosis: a non-fungal disease 30. Wanachiwanawin W, Thianprasit M, Fucharoen S, et al. mimicking orbital mycotic infections, with a retrospective review of the Fatal arteritis due to Pythium insidiosum infection in literature. Mycoses. 2004;47:14–23. patients with thalassaemia. Trans R Soc Trop Med Hyg. 1993;87: 27. Hotchkiss RD. A microchemical reaction resulting in the staining of 296–298. polysaccharide structures in ﬁxed tissue preparations. Arch Biochem. 1948; 31. Wanachiwanawin W, Mendoza L, Visuthisakchai S, et al. Efﬁcacy 16:131–141. of immunotherapy using antigens of Pythium insidiosum in the 28. Johansen DA. Microchemical methods. In: Plant Microtechnique. treatment of vascular pythiosis in humans. Vaccine. 2004;22: McGraw-Hill: New York; 1940:195. 3613–3621. 29. Chaiprasert A, Samerpitak K, Wanachiwanawin W, et al. Induction of 32. De Cock AW, Mendoza L, Padhye AA, et al. Pythium insidiosum sp. zoospore formation in Thai isolates of Pythium insidiosum. Mycoses. nov., the etiologic agent of pythiosis. J Clin Microbiol. 1987;25: 1990;33:317–323. 344–349.